The study team assembled genome-wide association summary statistics spanning four asthma subtypes, 12 neuropsychiatric disorders (NPDs), and six substance use behaviors to quantify shared genetic architecture. They reported that asthma shares a measurable genetic component with post-traumatic stress disorder (PTSD), major depressive disorder, and ADHD. The authors noted that smoking behavior did not fully account for these genetic correlations; in other words, the observed overlap persisted after considering smoking-related genetic effects. The analysis therefore suggests intrinsic, heritable links between respiratory and neuropsychiatric phenotypes rather than a simple causal chain mediated solely by tobacco exposure.
The narrative that emerged from the summary-statistic integration was not limited to pairwise correlations. The investigators applied a multi-trait GWAS framework designed to capture variants exerting pleiotropic influence across multiple phenotypes simultaneously. This approach prioritized loci that show concordant signals across the three trait domains—pulmonary, neuropsychiatric, and substance use—rather than signals unique to one condition.
From the multi-trait association results, the study team identified 354 independent genetic variants. They mapped those variants to 178 genes, which they described as pleiotropic because each gene associated with multiple trait categories in the analysis. The authors positioned these genes as representing the core shared genetic component among asthma, NPDs, and substance use behaviors.
To move from loci to likely cellular context, the investigators integrated single-cell transcriptomic data from blood, brain, and lung tissues. They reported pleiotropic gene overexpression across 57 cell types. Expression patterns were heterogeneous: some pleiotropic genes were enriched in immune cells circulating in blood and resident in lung tissue; others were preferentially expressed in central nervous system cell types, including microglia and specific neuronal populations. The single-cell work thus supplied a cell-type map linking the genetic signals to plausible biological substrates, and the authors used these expression patterns to propose mechanistic routes connecting respiratory and neuropsychiatric biology.
The investigators outlined three candidate mechanisms grounded in the pleiotropic gene set and its tissue-specific expression.
First, an immune-centered mechanism. Data indicated that many pleiotropic genes are expressed in microglia, peripheral blood immune cells, and lung-resident immune populations. The authors reported this pattern as consistent with shared immune dysregulation contributing both to airway inflammation in asthma and to neuroinflammatory processes implicated in several psychiatric conditions.
Second, a signaling-axis mechanism involving NOTCH pathway regulators. The authors identified EP300, RERE, and MAML3 as pleiotropic genes that modulate NOTCH signaling and that are expressed in neurons, microglia, and lymphoid progenitors. The reported co-expression across neuronal and immune progenitor compartments suggests a convergent role for NOTCH-related transcriptional control in both neural development/function and immune cell differentiation or activation.
Third, a ciliary-function mechanism. Several pleiotropic genes showed expression in ciliated cells of lung and brain. The study team proposed that defects in ciliary structure or function could plausibly affect mucociliary clearance in the airways and also influence neuronal/glial cellular processes in the brain, providing a potential shared cellular vulnerability linking asthma and some NPDs.
Those three mechanisms are presented as candidate biological routes emerging directly from cross-referencing pleiotropic loci with tissue- and cell-type expression; the authors did not report interventional or functional experiments validating causality within these pathways.
All datasets used in the study were publicly available GWAS summary statistics and scRNA-seq data, which the authors retrieved without access restrictions. The manuscript is a medRxiv preprint and has not undergone peer review; the authors explicitly state that the findings should not be used to guide clinical practice. They made the analysis code and data links available on GitHub, and they declared no competing interests.
The report focused on genetic mapping and transcriptional localization; it did not include clinical trial data, therapeutic dosing, or individual-level phenotype analyses. The authors did not report experimental validation of the candidate mechanisms; therefore, the functional and translational implications remain inferential. Finally, because the work relies on summary-level genomic data and publicly available single-cell atlases, the study team acknowledged the inherent limitations of those resources—heterogeneous cohort definitions, potential population stratification in source GWAS, and variable scRNA-seq sampling across tissues. Readers should interpret the proposed mechanisms as hypothesis-generating and wait for peer review and follow-up functional studies before considering clinical translation.